CharacterRange

Status: Stable

documented, exercised by the test suite and/or worked examples, with no known limitations recorded.

Description

CharacterRange["c1", "c2"]

Gives the list of characters with codes from that of c1 to that of c2.

CharacterRange[n1, n2]

Gives the characters with codes n1 through n2.

Examples (11)

Every input below was run against the current Mathilda build and its output recorded.

Basic examples (3)

In[1]:= CharacterRange["a", "e"]
Out[1]= {"a", "b", "c", "d", "e"}

In[2]:= CharacterRange[65, 70]
Out[2]= {"A", "B", "C", "D", "E", "F"}

In[3]:= Length[Alphabet[]]
Out[3]= 26

Applications (8)

In[4]:= CharacterRange["a", "e"]
Out[4]= {"a", "b", "c", "d", "e"}

Any Unicode characters, by code point

In[5]:= CharacterRange["α", "ε"]
Out[5]= {"α", "β", "γ", "δ", "ε"}

Integer endpoints are code points

In[6]:= CharacterRange[65, 70]
Out[6]= {"A", "B", "C", "D", "E", "F"}

Circled digits

In[7]:= CharacterRange[9312, 9316]
Out[7]= {"①", "②", "③", "④", "⑤"}

In[8]:= StringJoin[CharacterRange["0", "9"]]
Out[8]= "0123456789"

An empty range

In[9]:= CharacterRange["e", "a"]
Out[9]= {}

As a string pattern it is a character class

In[10]:= StringCases["x9a3b", CharacterRange["a", "z"]]
Out[10]= {"x", "a", "b"}

Mixed endpoint kinds: CharacterRange::argtype

In[11]:= CharacterRange["a", 70]
Out[11]= CharacterRange["a", 70]

Implementation notes

Algorithm. builtin_characterrange requires two arguments and converts each with range_end: a machine Integer that is a valid code point, or a String that decodes (via u8_decode) to exactly one code point. If either end is invalid, or the two ends are of different kinds (one string, one integer), it emits CharacterRange::argtype through mth_message and stays unevaluated. Otherwise char_list(lo, hi) encodes each code point from lo to hi with u8_encode into a one-character String, giving {} when lo > hi. As a string pattern, CharacterRange["a", "z"] is not expanded at all: string_pattern.c emits the PCRE character class [a-z] (put_class_char).

Data structures. A List of n = hi - lo + 1 one-character strings built by strings_char_expr. A surrogate code point inside an integer range (which u8_encode refuses) is rendered as "?" rather than producing invalid UTF-8.

Complexity / limits. O(n) time and memory in the size of the range; there is no lazy form, so a range spanning the whole code space allocates a million-element list. Ranges are by raw code point, with no collation or script awareness.

Attributes: Protected.

References

See also: Alphabet

• Source: src/strings/charcode.c
• Specification: docs/spec/builtins/string-operations.md
• Tests: tests/test_string_review.c
• Tests: tests/test_stringfns.c

Notes & additional examples

Notes

• The endpoints are both one-character strings or both non-negative integer codes; the range is by code point, so CharacterRange["ά", "ε"] also picks up the accented Greek letters lying between them.
• {} when the first code exceeds the second. Anything else emits CharacterRange::argtype and stays unevaluated.
• Inside a string pattern (StringCases, StringMatchQ, …) it is translated to a PCRE class [c1-c2] instead of being expanded.
• Attributes {Protected}.
• See also: Alphabet, FromCharacterCode, LetterCharacter.